Advances in AMSU NonSounding Products
Fuzhong Weng
R. Ferraro and N. Grody
NOAA/NESDIS/Office of Research and
Applications
Presented at the 12th International TOVS Studies Conference,
Lorne, Australia, 2/27, 2002
Outline
“
Describe a new suite of operational nonsounding at NOAA/NESDIS
AMSU Sensor
2. Product Overview & Retrieval Algorithms
3. Examples
4. Future/Data Availability
1.
NOAA AMSU Sensor
•Flown on NOAA-15 (May
1998) and NOAA-16 (Sept.
2000) satellites
•Contains 20 channels:
•AMSU-A
•15 channels
•23 – 89 GHz
•AMSU-B
•5 channels
•89 – 183 GHz
•6-hour temporal sampling:
•130, 730, 1330, 1930 LST
AMSU Observation Frequencies
Channel Frequency Channel
A1
A2
A3
A4
A5
A6
A7
Frequency
23.8 GHz
A8
55.5 GHz
31.4
A9-A14 57.290**
50.3
A15
89.0
52.8
B1
89.0
53.6
B2
150.0
54.4
B3-5
183.31**
54.9
AMSU-A and –B Scan Pattern
•Cross-track scan geometry
•2200 km swath width
•AMSU-A (30 FOV/scan; 48 km @ nadir)
•AMSU-B (90 FOV/scan; 16 km @ nadir)
AMSU and SSM/I Comparison
Parameter
Window Channels
Polarization
Scan Geometry
FOV’s
Swath Width
AMSU
SSM/I
23.8,31.4,50.3,89.0 (A) 19.4,22.2,37.0,85.5
89.0, 150.0 (B)
Mixed
V&H
0 - 48 deg
Fixed 45 deg
Vary with view angle:
45 (15) km/nadir
150 (50) km/limb
Vary with
frequency:
15 km @ 85 GHz
60 km @ 19 GHz
~2200 km
~1400 km
Microwave Surface & Precipitation
Products System (MSPPS) Product Suite
Product
Surface
AMSU-A
AMSU-B
Brightness Temperatures
Land/Ocean
z
Total Precipitable Water
Ocean
Cloud Liquid Water
Ocean
z
z
z
z
z
z
Rain Rate
Snow Cover
Sea Ice Concentration
Ice Water Path
Land/Ocean
Land
Ocean
Land/Ocean
z
z
z
Land Surface Emissivity
Land
z
Land Surface Temperature
Land
z
Moisture
Plumes
Limb Brightening
Emissivity
Variations
Cloud Features
Convective
Cells
“Dirty” Window
Enhanced
Precipitation
Signature
“Dirtier” Window
Oceanic TPW and CLW
L = a0[ln(Ts – TB31) – a1ln(Ts – TB23) - a2]
“ V = b0[ln(Ts – TB31) – b1ln(Ts – TB23)- b2]
“
Frontal
Bands
Convective
Clusters
Global Performance of Cloud
Retrievals (HKS Score)
The
The Weng2
Weng2 method
method is
is clearly
clearly more
more
skillful
skillful than
than Weng1
Weng1 or
or AAPP.
AAPP. The
The
Weng2
Weng2 method
method does,
does, however,
however,
show
show no
no advantage
advantage over
over Weng1
Weng1 in
in
the
the tropics.
tropics. The
The two
two 1D-vars
1D-vars show
show aa
similar
similar level
level of
of skill
skill to
to the
the Weng2
Weng2
method.
method. The
The Weng2
Weng2 method
method also
also
shows
shows less
less month-to-month
month-to-month variation
variation
in
in skill,
skill, especially
especially in
in the
the northern
northern
hemisphere
hemisphere (note
(note each
each month
month is
is
represented
represented by
by data
data from
from the
the 15th
15th
day
day of
of each
each month)
month)
Hansen-Kuiper
Hansen-Kuiper score
score == H-F
H-F
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
AAPP
Weng1
Weng2
1Dvar1
1Dvar2
Global
Extra tropics
Tropics
where
where H=Hit
H=Hit rate
rate and
and F=false
F=false alarm
alarm rate.
rate.
H
H == Number
Number where
where LWP
LWP >> 70
70 gm
gm-2-2 and
and validation
validation dataset
dataset == cloudy
cloudy divided
divided by
by total
total number
number
where
where validation
validation dataset=cloudy
dataset=cloudy
FF == Number
Number where
where LWP
LWP >> 70
70 gm
gm-2-2 and
and validation
validation dataset
dataset == clear
clear divided
divided by
by total
total number
number
where
where validation
validation dataset=clear
dataset=clear
Moisture
Plumes
Ice Water Path & Rain Rate
’Physical retrieval of ice water path (IWP) and
particle size (De) using AMSU-B 89 and 150 GHz:
•De ~ Ω(89)/Ω(150)
Zhao and Weng ( 2002, JAM)
•IWP ~ De*(Ω/Ω(89,150))
’Assumptions made on size-distribution & density
’IWP to rain rate based on limited cloud model
data and comparisons with in situ data:RR = A0 +
A1*IWP + A2*IWP2
’Effects of surface misidentification (desert &
snow) reduced using 89 and 150 GHz
Ice Water Path & Rain Rate
130 am
730 am
130 pm
730 pm
Morning Maximum –
Weak Diurnal Cycle
130 am
730 am
130 pm
730 pm
Strong Diurnal Cycle –
Morning Min
Evening Max
AMSU-B vs NCEP (BINNED DATA)
AMSU-B
JUNE5-8-01
T. S. ALLISON
25
20
15
10
5
0
y = 0.3518x + 4.1592
R = 0.69
0
5
10
15
NCEP
20
25
Snow Cover
“
ω31 = TB 23 - TB 31 - 2.0 (best for refrozen snow)
“
ω89= TB 23 - TB 89 - 3.0 (best for new/shallow snow)
“
If ω31 < 3 and TB 23 < 215 K (glacial snow)
There are also checks to eliminate false signatures of
snow due to precipitation and cold deserts.
“
130 pm
Sea Ice Concentration
“ 23 = a + b TB23 + c TB31+ d TB50, where
coefficients are function of µ=cos(θ)
“ water= 0.1824 + 0.9048 µ - 0.6221 µ2
{
“ ice=
“
0.93
if (TB23 – TB31) < 5 K
first-year
0.87 if 5 < (TB23 – TB31) < 10 K
multi-year
0.83 if 10 < (TB23 – TB31)
Sea Ice (%) =100 (e - ewater) / (eice - ewater)
AMSU Sea Ice
11/04/01
Land Surface Emissivity
“ derived for 23, 31 and 50 GHz
“ i= b0, i + b 1, i TB23 + b 2, i TB223
+ b 3, i TB31 + b 4, i TB231
+ b 5, i TB50 + b 6, i TB250
where i=1,2,3
“ Coefficients derived through regression
relationships between “truth” emissivity and
brightness temperatures
“ The truth is defined by removing the atmospheric
effects.
Land Surface Temperature
“ Regression
relationship between
TB’s and NWP global model fields
“ Tsfc = f(TB23, TB31, TB50, µ)
“ Includes linear and non-linear terms
Hurricane Temperature & Wind
Hurricane Bonnie
Balanced Wind Retrieval
Retrieve temperatures
from AMSU
“ Derive geopotential
using hydrostatic
equation
“ Solve the balance
equation for
streamfunction
“ Calculate wind vectors
from streamfunction
Empirically define V0(r)
Require: PCEN, RMAX and VMAX.
“
V=f(p)V0(r)+V’
V=kx∇ψ
Balance Equation:
f∇ ψ + 2(ψ xxψ yy − ψ xy 2 ) + ψ x f x + ψ y f y = ∇ 2φ
2
Φ
Hydrostatic relation:
1 ∂φ
T =−
R ∂ ln p
T
Satellite retrieved T
Specify PSFC
Require: PCEN,
RMAX and VMAX
AMSU Derived Balanced Winds
During the Mature Stage
AMSU Derived Balanced Winds
During the Incipient Stage
NCEP Eta Model Winds
During the Incipient Stage
AMSU Derived Balanced Wind vs.
GPS Dropsonde Measurements
Vertical Motion Derived from
AMSU Rain Rate
AMSU rain rate is used to
derive the latent heat
profile
Latent heat is used to
solve vertical velocity by
solving the omega
equation
Continuity equation is
used to derive the
divergent wind.
Shown are composite
balanced and divergence
and vertical motion of
hurricane Bonnie
Future Plans
Continue operational generation of AMSU
nonsounding products through NOAA-M,N,N’
“ Algorithm improvements
“
“
“
“
New algorithms being considered
“
“
“
Improved physical models
Better sensor characterization
Snow depth
Soil wetness (temporal information)
Synergy with NWP assimilation
References
“
“
“
“
“
Ferraro, R.R, F. Weng, N.C. Grody and L. Zhao, 2000:
Precipitation characteristics over land from the NOAA-15
AMSU Sensor. Geophy. Res. Let., 27, 2669-2672.
Weng, F., B. Yan and N. Grody, 2001: A microwave land
emissivity model. J. Geophys. Res., 106, 20,115-20,123.
Zhao, L. and F. Weng, 2002: Retrieval of ice cloud
properties using the Advanced Microwave Sounding Unit
(AMSU). J. Appl. Meteor., 41, 384-395.
Zhu, T., D. Zhang, and F. Weng, 2002: A hurricane
model initialization scheme using AMSU, MWR
(accepted).
Weng, F. et al, 2002: AMSU cloud and precipitation
algorithms. Radio Science (submitted).
Data Availability & Information
“
Our web site:
http://orbit-net.nesdis.noaa.gov/arad2/microwave.html
“
Data formats:
“
“
“
“
NESDIS/OSDPD:
“
“
“
HDF-EOS
BUFR (~2002)
McIDAS (AWIPS ~2002)
Operational Request
SAA (~2002)
NESDIS/ORA:
“
Weng/Ferraro/Grody